BEGIN:VCALENDAR
VERSION:2.0
PRODID:Linklings LLC
BEGIN:VTIMEZONE
TZID:Asia/Hong_Kong
X-LIC-LOCATION:Asia/Hong_Kong
BEGIN:STANDARD
TZOFFSETFROM:+0800
TZOFFSETTO:+0800
TZNAME:HKT
DTSTART:19911015T033000
END:STANDARD
END:VTIMEZONE
BEGIN:VEVENT
DTSTAMP:20251218T030657Z
LOCATION:Meeting Room S421\, Level 4
DTSTART;TZID=Asia/Hong_Kong:20251215T132000
DTEND;TZID=Asia/Hong_Kong:20251215T133100
UID:siggraphasia_SIGGRAPH Asia 2025_sess104_papers_2088@linklings.com
SUMMARY:Topology-Aware Optimization of Gaussian Primitives for Human-Centr
 ic Volumetric Videos
DESCRIPTION:Yuheng Jiang (Max Planck Institute for Informatics, Saarland I
 nformatics Campus; ShanghaiTech University); Chengcheng Guo, Yize Wu, Yu H
 ong, Shengkun Zhu, and Zhehao Shen (ShanghaiTech University); Yingliang Zh
 ang (DGene Inc.); Shaohui Jiao and Zhuo Su (ByteDance Inc.); Lan Xu (Shang
 haiTech University); and Marc Habermann and Christian Theobalt (Max Planck
  Institute for Informatics, Saarland Informatics Campus)\n\nVolumetric vid
 eo is emerging as a key medium for digitizing the dynamic physical world, 
 creating the virtual environments with six degrees of freedom to deliver i
 mmersive user experiences. However, robustly modeling general dynamic scen
 es, especially those involving topological changes while maintaining long-
 term tracking remains a fundamental challenge. In this paper, we present T
 aoGS, a novel topology-aware dynamic Gaussian representation that disentan
 gles motion and appearance to support, both, long-range tracking and topol
 ogical adaptation. We represent scene motion with a sparse set of motion G
 aussians, which are continuously updated by a spatio-temporal tracker and 
 photometric cues that detect structural variations across frames. To captu
 re fine-grained texture, each motion Gaussian anchors and dynamically acti
 vates a set of local appearance Gaussians, which are non-rigidly warped to
  the current frame to provide strong initialization and significantly redu
 ce training time. This activation mechanism enables efficient modeling of 
 detailed textures and maintains temporal coherence, allowing high-fidelity
  rendering even under challenging scenarios such as changing clothes. To e
 nable seamless integration into codec-based volumetric formats, we introdu
 ce a global Gaussian Lookup Table that records the lifespan of each Gaussi
 an and organizes attributes into a lifespan-aware 2D layout. This structur
 e aligns naturally with standard video codecs and supports up to 40× compr
 ession.TaoGS provides a unified, adaptive solution for scalable volumetric
  video under topological variation, capturing moments where ``elegance in 
 motion'' and ``Power in Stillness''— delivering immersive experiences that
  harmonize with the physical world.\n\nRegistration Category: Full Access,
  Full Access Supporter\n\nSession Chair: Xuejin Chen (University of Scienc
 e and Technology of China)\n\n
END:VEVENT
END:VCALENDAR